Colonization rather than fragmentation explains the geographical distribution and diversification of treefrogs endemic to Brazilian shield sky islands

• Published in: Journal of biogeography 2022-04, Vol.49 (4), p.682-698
• Main Authors: Oswald, Caroline Batistim, Lemes, Priscila, Thomé, Maria Tereza C., Pezzuti, Tiago Leite, Santos, Fabricio Rodrigues, Garcia, Paulo Christiano de Anchietta, Leite, Felipe Sá Fortes, Magalhães, Rafael Félix
• Format: Article
• Language: English
• Subjects: Abiotic factors; approximate Bayesian computation; Bayesian analysis; Biodiversity; Biota; Bokermannohyla; Bokermannohyla saxicola; campo rupestre; Colonization; Current distribution; Deoxyribonucleic acid; Dispersal; Dispersion; Divergence; DNA; ecological niche modelling; Ecological niches; Endemism; Espinhaço Range; Evolution; flickering connectivity system; Fragmentation; Frogs; Genetic diversity; Geographical distribution; Habitat fragmentation; Mountains; OCBIL; Phylogeny; Pleistocene; Pliocene; Species; species delimitation
• ISSN: 0305-0270; 1365-2699
• DOI: 10.1111/jbi.14320

Aim Geographical patterns of montane biodiversity worldwide are related to biotic and abiotic factors, such as historical climate dynamics and species dispersal capabilities, which affect the biota from population to community levels. Understanding of processes related to population diversification in extra‐Andean Neotropical mountains remains largely unknown. Here, we tested how colonization of new areas as opposed to fragmentation of geographical ranges influenced the distribution and diversification of frogs from Brazilian mountains. Location Espinhaço Range, the largest extra‐Andean mountain range in South America. Taxon Ledge treefrog Bokermannohyla saxicola. Methods We used multilocus DNA and morphometric data throughout the species' distribution to delimit lineages, infer phylogenetic relationships and estimate divergence times. We used ecological niche modelling (ENM) and approximate Bayesian computation (ABC) to reconstruct changes of geographical distribution and population sizes for testing alternative hypotheses of diversification concerning fragmentation and colonization. Results We found four evolutionary significant lineages that diverged from the Pliocene to the Early Pleistocene. All lineages were validated by DNA and morphometric data by independent and joint analyses. ENM showed that climatic fluctuations might have influenced the species' distribution, whereas ABC model selection further supports associated demographic changes. Taken together, a Pleistocene jump‐dispersal scenario best explains the diversification of the Bokermannohyla saxicola lineages. Main conclusions Endemism in the Espinhaço Range is usually explained by ancient diversification, associated with long‐term isolation of climatic stable areas following climate‐driven habitat fragmentation. Our results challenge this general view, indicating recent diversification, habitat expansion and colonization of new areas as important processes explaining the current distribution and genetic diversity of B. saxicola.